run-22/checkpoint-1290/trainer_state.json

{
  "best_metric": 0.8997673516867003,
  "best_model_checkpoint": "tiny-bert-sst2-distilled/run-22/checkpoint-1290",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1290,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.1603200435638428,
      "learning_rate": 0.00015804016249699036,
      "loss": 0.3253,
      "step": 645
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8945327646374563,
      "eval_f1": 0.19047619047619047,
      "eval_loss": 0.29357558488845825,
      "eval_mcc": 0.23285387927696083,
      "eval_precision": 0.6213592233009708,
      "eval_recall": 0.11247803163444639,
      "eval_runtime": 9.9704,
      "eval_samples_per_second": 517.33,
      "eval_steps_per_second": 16.248,
      "step": 645
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3251060247421265,
      "learning_rate": 0.00011853012187274275,
      "loss": 0.2915,
      "step": 1290
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8997673516867003,
      "eval_f1": 0.2385861561119293,
      "eval_loss": 0.2915143072605133,
      "eval_mcc": 0.2949928981799802,
      "eval_precision": 0.7363636363636363,
      "eval_recall": 0.14235500878734622,
      "eval_runtime": 9.9347,
      "eval_samples_per_second": 519.192,
      "eval_steps_per_second": 16.307,
      "step": 1290
    }
  ],
  "logging_steps": 500,
  "max_steps": 3225,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 3173885065200.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.9879592289455201,
    "learning_rate": 0.00019755020312123793,
    "num_train_epochs": 5,
    "temperature": 26
  }
}